Tracing pain pathways from stimulus to report
Lauren Y.
1
Atlas ,
Matthew
1
Davidson ,
Niall
1
Bolger ,
Kate
1
Dahl ,
Martin
2
Lindquist ,
Tor D.
1
Wager
1Columbia University Department of Psychology, 2Columbia University Department of Statistics
Columbia Psychology SCAN group
http://www.scan.psych.columbia.edu/
RESULTS
INTRODUCTION
Noxious stimulus intensity and reported pain are highly correlated.
However, there is rarely a perfect mapping between applied
nociceptive stimulation and perceived pain.
Studies have revealed differences in activity linearly related to
temperature and activity linearly related to reported pain (Craig et
al., 2000) using separate linear regressions.
Temperature regression: L. posterior insula, R. anterior
insula
Reported pain regression: R. anterior insula, OFC
Path A: Which brain regions
show greater activity for higher
levels of thermal stimulation?
STIMULUS-RELATED REGIONS
RESULTS
Path B: For a given level of heat,
which brain regions predict
magnitude of pain ratings?
RESPONSE-RELATED REGIONS
Regression path model
Studies have not traditionally examined brain regions involved in the
pain reporting process controlling for activity related to nociceptive
input.
It is likely that specific brain regions play an important role in
constructing the pain experience from a given level of stimulation.
These mediator regions may be identified using whole brain multilevel mediation analyses, to localize pathways that link stimulus
intensity to reported pain experience.
Direct Effect
METHODS
•Pain calibration and experimental design
•20 subjects (mean age = 30, 8 females)
•Thermal stimulation delivered to 3 sites on left forearm with 16mm thermode
(Medoc, Inc.))
Pain Rating
Positively related: Bilateral SII, contralateral
SI, dACC, bilateral thalamus, PAG, midbrain, R
ventral striatum, cerebellum, bilateral anterior
insula, bilateral posterior insula, RVLPFC,
posterior cingulate
Positively related: Contralateral SII, dACC,
bilateral thalamus, bilateral putamen, cerebellum,
bilateral anterior insula
Inversely related: Occipital lobe, R inferior
Inversely related: mOFC, R DLPFC, mPFC,
temporal gyrus
occipital lobe, post. cingulate cortex, bilateral
parahippocampal gyrus
Warm
Low
Med
PATH C / C’
Applied Heat
High
Perceived Pain
2. Estimate trial-by-trial height,
1. Fit basis functions trial-by-trial width, delay, and area under the
curve (AUC)
Mediation analyses also identified stimulus-related regions and
response-related regions. Many key components of the traditional
‘pain matrix’ were identified in these path analyses.
Brain Mediators
L. Precentral gyrus
Post. Insula
Ant. Insula
L. IPL
SII
Regions identified in the mediation analysis were then included in a
linear path model illustrating the role of these regions in the
pathway from noxious stimulation to the pain experience.
R. Precentral gyrus
R. Occipital
Ant. Insula
Ventral
striatum
(bilateral)
R
3. Use trial level parameters in multi-level
mediation
L (Ipsilateral)
SMA
•Whole-brain multi-level mediation
Three linear equations:
1. y = cx + ey
2. m = ax + em
3. y = bm + c'x + e'y
L. VLPFC
Future analyses will examine potential moderators of these
pathways, including expectancy-related brain activity, individual
difference measures, and cognitive control regions,
R (Contralateral)
Precuneus
dACC
dACC
A test for mediation should
satisfy the following criteria:
1. M should be related to X (a
effect)
2. M should be related to Y
after controlling for X (b
effect)
3. The indirect relationship
(a*b)should be significant
SUMMARY
Multi-level mediation analyses identified a network of brain regions
that explained a significant amount of the covariation between
temperature and pain reports. Key mediators included right anterior
insula, dorsal anterior cingulate cortex, and left cerebellum.
A*B: Which brain regions
contribute to the relationship
between applied heat and
perceived pain?
•Voxelwise single trial analysis
Data from mediators, stimulus-related regions, and responserelated regions were then extracted. Stimulus and rating vectors
were included in a multiple regression model predicting activity for
each region, and coefficients were sorted in order to identify the role
of each region in the continuum from noxious stimulus to perceived
pain.
pgACC
REFERENCES
Craig AD, Chen K, Bandy D, Reiman EM (2000). Nat Neurosci 3:184-190.
Thalamus
p<.001 one-tailed, 5 contiguous voxels
p<.01, 20 contiguous voxels
Cerebellum
Thalamus
CONJUNCTION ANALYSIS: Right
anterior insula, dorsal ACC, and left
cerebellum were significant for all
three paths (A, B, A*B).
For more information on whole-brain multi-level mediation and
moderation analyses, see poster B95 (tomorrow morning):The M3
Toolbox: the Multi-level Mediation/Moderation Framework for
Connectivity Analyses in fMRI Data
Matthew Davidson, Lauren Atlas, Martin Lindquist, Niall Bolger & Tor
Wager
Poster reprints at
http://www.columbia.edu/cu/psychology/tor/
Contact:
laurenatlas@gmail.com